1. Field of the Invention
The present invention relates to imaging arrays and systems. More particularly, the present invention relates to a single-chip high-sensitivity imaging system for producing serial-scanned outputs.
2. The Prior Art
Imaging circuits typically include a two-dimensional array of photosensors, each of which converts a pixel of light energy from an image into an electrical signal representing the light value of the pixel. The imaging circuit scans the photosensors to read out the electrical signals representing the light values of all pixels from an image.
As a result of the advances in semiconductor technology, single chip imaging circuits are known in the art. The most common single-chip imaging technology is the charge coupled device (CCD) camera. A CCD device operates by accumulating photo-generated charge in potential wells in a semiconductor substrate at or slightly under the surface of the substrate. The depth of these wells is controlled by the voltage on gate electrodes located just above the surface of the substrate. By manipulating the voltage on several of the gates, the accumulated charge can be shifted along the surface to a sensing point, where it is amplified into a signal which can be driven off of the chip.
Using modern MOS processing techniques, the transport of charge in a CCD structure can be accomplished with almost perfect efficiency at video rates; typically less than ten parts per million of the charge in a well is lost in each transfer. In a typical serial-scanned CCD imager, such as is used for video applications, the charge is shifted out once per frame, and hence the total charge in each well is collected over the entire frame time, typically 1/60 second. Sensing technologies which operate in this manner, by accumulating charge over the entire time between scanout events, are called integrating photosensors.
Despite its technical achievements and wide spread commercial success, CCD technology has its limitations. Because the photo-generated charge is shifted directly, the gain of the device (electrons out per photon in) is always less than unity. Because the charge stored in each well is limited, the CCD imager has severe dynamic-range constraints. At the low end, the sensitivity is limited by how small a charge can be sensed above the noise of the readout amplifier. At the high end, the range is limited by the total charge that can be shifted from one well into the next.
To overcome these dynamic-range limitations of the CCD imager, a number of attempts have been made to use phototransistors to sense incoming light, such as seen in Analog VLSI and Neural Systems, by Carver A. Mead, Addison Wesley Publishing Co., 1989, at pp. 260-261. Phototransistors can have gain of over 100 electrons per absorbed photon. In a typical application, the photocurrent from each phototransistor is fed into a device with exponential current-voltage characteristics, such as a pair of diode-connected MOS transistors. The output voltage of such a device is then a logarithmic function of the light level. These devices show much wider dynamic range than do CCD devices, but are not as sensitive at low light levels because they operate in instantaneous-current mode rather than in integrating mode.
Photons absorbed in the neighborhood of a phototransistor collector-base or emitter-base junction create electron-hole pairs that are collected by the nearest p-n junction. Minority carriers collected by either junction act as base current, and are multiplied by the current gain of the transistor to produce the collector current. The emitter current is the sum of the base current and the collector current. For this reason, the emitter current is usually used as the output of the sensor.
Prior art phototransistor sensors suffer from low-current limitations. The current-gain of the phototransistor sensor device is high at high photocurrents, where it is not needed, and is low at low photocurrents, where it would desirably be higher. Because of this behavior of phototransistors, they have not found widespread use in commercial scanned imagers.
It is an object of the present invention to provide an improved single-chip serial-scanned imager, which can exhibit higher sensitivity at lower light levels than current mode imagers, and in the same device operate over a dynamic range larger than that achievable with CCD imagers.